Agentic and Autonomous Data Management
By 2027, 40% of all OT data will be integrated into platforms and apps autonomously, thanks to the use of AI agents created specifically for certain data. This will push manufacturers to move away from centralized data models.
As manufacturers adopt agentic AI and federated data architectures, executives must reassess risk management and governance.
IDC found that the most important areas of risk to mitigate across multiple locations and with partners across your ecosystems are:
- Cybersecurity of collaboration platforms
- Environmental sustainability
- Supply chain execution and logistics
- Product/service quality issues
- Regulatory compliance changes
Hojlo points to new approaches that help manufacturers confirm the accuracy of both external data and AI-generated insights by validating results against trusted internal data. Overcoming skepticism and building trust in AI guidance remains critical.
WATCH: Discover the security issues demanding attention in 2026.
Human-Robot Collaboration and Workforce Transformation
IDC predicts that AI will reshape manufacturing workforces through continuous human-robot learning and personalized training, reducing downtime and accelerating skill development.
Workforce strategy is now inseparable from AI strategy. Within the next few years, AI agents will be embedded in a significant share of manufacturing roles. “Manufacturers that can capture this tacit knowledge as part of their overall supporting, synchronous data model will have an advantage,” Hojlo says.
Downtime considerations further elevate the importance of simulation. “There cannot be downtime in production, as this is too expensive,” Hojlo says, driving increased use of “simulation and digital twins to model the use of new technologies and decisions before they are made.”
EXPLORE: How experts predict workplace technology will evolve in 2026.
Security, Resilience and Ecosystem Collaboration
As AI embeds itself in OT cybersecurity, validation becomes critical. By 2029, IDC says, 75% of large manufacturers will use AI-powered cyber defense to detect threats faster and with less manual effort.
Hojlo says manufacturers must verify and validate AI-driven decisions using formal quality assurance processes, rather than relying solely on AI model outputs. Meanwhile, ransomware continues to pose a significant risk, frequently exploiting legacy systems and phishing attacks.
